early maternal employment and family wellbeing

NBER WORKING PAPER SERIES

EARLY MATERNAL EMPLOYMENT AND FAMILY WELLBEING

Pinka ChatterjiSara Markowitz

Jeanne Brooks-Gunn

Working Paper 17212http://www.nber.org/papers/w17212

NATIONAL BUREAU OF ECONOMIC RESEARCH1050 Massachusetts Avenue

Cambridge, MA 02138July 2011

The authors gratefully acknowledge that the project was supported by Grant Number R03HD052583from the National Institute of Child Health And Human Development. The content is solely the responsibilityof the authors and does not necessarily represent the official views of the National Institute of ChildHealth And Human Development, the National Institutes of Health, or the National Bureau of EconomicResearch.

© 2011 by Pinka Chatterji, Sara Markowitz, and Jeanne Brooks-Gunn. All rights reserved. Short sectionsof text, not to exceed two paragraphs, may be quoted without explicit permission provided that fullcredit, including © notice, is given to the source.

Early Maternal Employment and Family WellbeingPinka Chatterji, Sara Markowitz, and Jeanne Brooks-GunnNBER Working Paper No. 17212July 2011JEL No. I1

ABSTRACT

This study uses longitudinal data from the NICHD Study on Early Child Care (SECC) to examinethe effects of maternal employment on family well-being, measured by maternal mental and overallhealth, parenting stress, and parenting quality. First, we estimate the effects of maternal employmenton these outcomes measured when children are 6 months old. Next, we use dynamic panel data modelsto examine the effects of maternal employment on family outcomes during the first 4.5 years of children’slives. Among mothers of six month old infants, maternal work hours are positively associated withdepressive symptoms and self-reported parenting stress, and negatively associated with self-rated overallhealth among mothers. Compared to mothers who are on leave 3 months after childbirth, motherswho are working full-time score 22 percent higher on the CES-D scale of depressive symptoms. However,maternal employment is not associated with the quality of parenting at 6 months, based on trainedassessors’ observations of maternal sensitivity. Moreover, during the first 4.5 years of life as a whole,we find only weak evidence that maternal work hours are associated with maternal health, and no evidencethat maternal employment is associated with parenting stress and quality. We find that unobservedheterogeneity is an important factor in modeling family outcomes.

Pinka ChatterjiState University of New York at AlbanyEconomics Department1400 Washington AvenueAlbany, NY 12222and [email protected]

Sara MarkowitzDepartment of EconomicsEmory University1602 Fishburne Dr.Atlanta, GA 30322and [email protected]

Jeanne Brooks-GunnColumbia UniversityNational Center for Children and Families525 West 120th Street, Box 39New York, NY [email protected]

3

I. Introduction Maternal employment has been the norm in the US since the 1980’s. As of 2008, 71

percent of mothers of children under age 18 participated in the labor force (US BLS, 2009).

Rates of labor force participation in 2008 were somewhat lower, but still high, for mothers of

young children and infants – 64 percent of mothers with children under 6 years old, and 56

percent of mothers of infants participated in the labor force in 2008 (US BLS, 2009). Child-

rearing and market work are both time-intensive activities. Thus, there has been concern that

maternal employment harms children by reducing the quantity and quality of time mothers spend

with their families (Baum 2003; Ruhm, 2004; Cawley & Liu, 2007).

When mothers reallocate their time from home to market work, however, this shift

potentially affects not just the health and wellbeing of children but also the health and well-being

of the parents and the family as a whole (Bianchi 2000; Riggio, 2006). In the economics

literature, previous research on the effects of maternal employment has focused on a narrow set

of child outcomes - scores on cognitive tests and a single behavioral assessment, the Behavior

Problems Index (BPI). However, because only child outcomes and not maternal or family

outcomes are examined, we obtain an incomplete picture of the effects of maternal employment

from these studies. To develop public policies that meet the needs of a society in which most

mothers are employed, we need a broader knowledge base regarding how maternal employment

affects families.

This study uses longitudinal data from Phases I and II of the NICHD Study on Early

Child Care (SECC) to examine the effects of maternal employment on family well-being,

measured by maternal mental and overall health, parenting stress, and parenting quality. First, we

estimate the effects of maternal employment on these outcomes measured when children are 6

4

months old, a point at which child-rearing is particularly time-intensive and maternal

employment may have its most important effects. Next, we take advantage of the longitudinal

aspect of the SECC and use dynamic panel data models to examine the effects of maternal

employment on family outcomes during the first 4.5 years of children’s lives. Notably, these

models account for both unobserved heterogeneity as well as state dependence in the maternal

health and parenting outcomes. The primary contributions of this paper are: (1) we examine the

effects of maternal employment on a broader set of family outcomes than most US-based studies

have considered, including parenting quality assessed in a laboratory setting; (2) in addition to

work hours, we consider effects of work characteristics such as job flexibility; (3) we draw on

longitudinal data to gauge whether effects persist over early childhood; and (4) we use empirical

methods that address the potential endogeneity of maternal work hours, as well as the dynamic

nature of the relationship between maternal employment and family outcomes.

Our findings indicate that maternal employment is associated with reductions in family

wellbeing when children are 6 months old. Among mothers of six month old infants, maternal

work hours are positively associated with depressive symptoms and self-reported parenting

stress, and negatively associated with self-rated overall health among mothers. Compared to

mothers who are on leave 3 months after childbirth, mothers who are working full-time score 22

percent higher on the CES-D scale of depressive symptoms. However, maternal employment is

not associated with the quality of parenting at 6 months, based on trained assessors’ observations

of maternal sensitivity. Moreover, during the first 4.5 years of life as a whole, we find only weak

evidence that maternal work hours are associated with maternal health, and no evidence that

maternal employment is associated with parenting stress and quality. We find that unobserved

heterogeneity is an important factor in modeling family outcomes.

5

II. Effects of early maternal employment on family outcomes

There is an extensive theoretical and empirical literature outside of economics that

highlights the importance of the family environment, parenting, and maternal health in shaping

children’s health and developmental trajectories (National Research Council and Institute of

Medicine, 2000; Belsky, 1988; Coleman,1988; Bornstein, 2002). Parenting behaviors, such as

nurturance, discipline, and teaching, as well as the home environment have powerful influence

on children’s wellbeing and development (Brooks-Gunn & Markman, 2005). Parenting and the

home environment have been linked to cognitive and behavioral outcomes such as children’s

academic achievement, social functioning, and health (Bornstein, 2002; Collins et al., 2000;

Steinberg, 2001; Steinberg & Sheffield-Morris, 2001). Specifically, parental hostility, low

nurturance, parenting stress, physical discipline and other harsh parenting practices are

associated with aggression, low self-control, higher levels of externalizing behavior problems,

and other mental health problems in children (Feldman et al., 2000; Qi & Kaiser, 2003; Barry et

al., 2005; Ispa et al. forthcoming).

Poor parental health also reduces the quality of time mothers spend with their children

and is associated with adverse outcomes. Numerous studies show that clinical depression in

mothers as well as self-reported depressive symptoms, anxiety, and psychological distress, are

important risk factors for adverse emotional and cognitive outcomes in their children,

particularly during the first few years of life (Gray et al. 2004; NICHD Early Child Care

Research Network, 1999; Petterson & Albers, 2001). Depressed mothers of infants are less

interactive with and less responsive to their children (Campbell et al. 1995), and are less likely to

seek appropriate health care for their children (Minkovitz et al., 2005). Compared to infants of

6

healthy mothers, infants of depressed mothers are more negative and less playful (Cohn et al.,

1986; Field, 1984), have more behavior problems during childhood (Field 1984; Barry et al.,

2005; Essex et al., 2001; Lyons-Ruth et al., 1997; Hay et al., 2003), and they are more likely to

eventually develop psychopathology during childhood and adulthood (Downey & Coyne, 1990;

Kim-Cohen et al., 2005).

Despite the importance of maternal health and parenting outcomes, there has been little

attention in the economics literature to the effects of maternal employment on outcomes of

family members other than children. Most research focuses on the effects of maternal

employment on children’s academic and behavioral outcomes. Recent research indicates that

early maternal employment increases the frequency of child behavior problems, and detracts

from school readiness, verbal ability, and test scores (Berger et al., 2008; Berger et al., 2005;

Brooks-Gunn et al., 2002; Hill et al., 2005; Waldfogel et al., 2002; Waldfogel, 2002; Ruhm,

2004; Ruhm 2008; Baum, 2003; James-Burdumy, 2005; Gregg et al., 2005 ). Full-time

employment during the first eighteen months is particularly harmful for children’s cognitive and

behavioral outcomes (Gregg et al., 2005; Baum, 2003; Brooks-Gunn et al., 2002).

These effects of maternal employment on children vary by the characteristics of children

(e.g., age, race/ethnicity, and gender of child) and families (e.g., SES), and also by child care

quality (e.g., type of child care arrangement). The negative effects of maternal employment tend

to be stronger for children from more advantaged, majority-race backgrounds (Ruhm, 2008;

Berger et al., 2008; Gregg et al., 2005). Moreover, non-standard maternal work schedules during

the child’s first few years are associated with child behavior problems (Daniel et al., 2009; Han,

2005), and recent research based on the SECC shows that more hours of non-relative child care

are associated with child behavior problems up to age 15 (Vandell et al., 2010).

7

A few recent studies focus on the effect of one aspect of maternal employment – the

length of maternity leave – on maternal health, which is one of the measures of family wellbeing

we consider in the present study. These studies offer mixed evidence that maternity leave is

associated with maternal health. Based on Canadian data, Baker and Milligan (2008) evaluate a

mandated increase in the number of weeks of maternity leave granted to new parents. They find

that increasing paid leave benefits from a maximum of 25 weeks to 50 weeks has no influence on

maternal health measured by self reported health status, a depression scale, an indicator of post

partum depression and a count of post-partum physical problems. In the US context, Chatterji &

Markowitz (2005, 2008) use data from the 1988 National Maternal and Infant Health Survey

(NMIHS) and the Early Childhood Longitudinal Study – Birth Cohort (ECLS-B) to examine the

association between maternity leave length and maternal health. The findings from these two

papers suggest that longer maternity leave (paid and un-paid) is associated with lower levels of

maternal depressive symptoms, a lower likelihood of the mother having frequent outpatient visits

during the first six months after childbirth, and better self-reported overall health among

mothers.

To our knowledge, only one recent study in economics has focused on the effects of

maternal employment on the wellbeing of the family. Baker, Gruber & Milligan (2008) take

advantage of a natural experiment in which one Canadian province (Quebec) introduced a

comprehensive, highly subsidized child care system. This policy change led to a rise in child

care usage, an increase in maternal employment, and an increase in children’s adverse health and

developmental outcomes in Quebec relative to the rest of Canada. These negative effects on

children are consistent with the US-based literature on the effects of maternal employment on

child outcomes. Baker, Gruber & Milligan, however, also examine family outcomes which have

8

not been studied in the US context. These authors find that the policy change was associated

with less effective parenting, less satisfaction with marital relationships, increases in maternal

depressive symptoms, and decline in the overall self-assessed health of fathers (but not mothers).

This study is notable in that the authors examine a range of family indicators of wellbeing, not

just children’s outcomes, and all effects suggest detrimental effects of child care and maternal

employment on families.

We build on this recent paper by examining these types of effects in the US context

where, compared to Canada and other industrialized countries, mothers return to work early and

at best have limited access to paid leave and affordable, high quality child care. This paper

contributes to the growing economics literature on the effects of maternal employment in three

respects. First, it expands the range of outcomes considered by examining the effects of

maternal employment on several measures of family wellbeing that are critical to young

children’s development -- maternal mental health, maternal overall health, parenting stress, and

the quality of parenting.

Second, we use data from the SECC, the only national, longitudinal data set available that

includes detailed information on maternal work hours as well as state-of-the-art measurement of

family outcomes, including laboratory assessments of parenting behaviors. Using the SECC, we

attempt to address the possibility that unmeasured factors confound an observed association

between maternal employment and family outcomes by estimating models that include an

unusually extensive set of controls, including measures of initial family wellbeing and maternal

ability, and by estimating dynamic panel models which address unobserved heterogeneity as well

as state dependence in the outcomes.

9

Third, we build on prior work by examining not only the effects of maternal work hours

on family wellbeing, but also the effects of the characteristics of her work – specifically, whether

the mother does some work at home, whether she has flexible hours, whether she travels

overnight for work, and whether she works non-standard hours. By examining these

characteristics in addition to hours of work, we gain a richer and more nuanced picture of how

maternal employment affects families.

III. The NICHD Study of Early Child Care (SECC)

The SECC is a longitudinal study designed to examine the relationship between child

care and children’s development. In 1991, the SECC enrolled 1,364 healthy infants with

English-speaking, adult mothers from ten sites across the United States.1 The sampling plan

ensured representation of mothers who planned to work or attend school full-time, as well as

mothers who planned to work/attend school part-time and mothers who planned to be home full-

time with their infants (see NICHD Early Child Care Research Network, 1994, for a detailed

description of the study design). The SECC was conducted in three phases, following children

and their families from birth until age 15. This paper is based on data from Phase I and Phase II

of the study, which includes repeated assessments of children, parents, and caregivers, as well as

multiple home and telephone interviews with parents, caregivers, and child care directors. We

draw on data collected from the time children were 1 month old until children were 54 months or

about 4.5 years old.

Compared to the NLSY79 and other similar surveys used in prior work, measurement of

maternal and child outcome variables is much more extensive in SECC. Data were collected

1 Little Rock, AR; Irvine, CA; Lawrence, KS; Boston, MA; Philadelphia, PA; Pittsburgh, PA; Charlottesville, VA; Morganton, NC; Seattle, WA; Madison, WI

10

from children and their families in a variety of settings, including the child’s home, the child care

setting (if used), a laboratory playroom, and through telephone contacts with a parent. Outcomes

are measured using a variety of methods including use of trained observers, interviewers,

questionnaires, and testing (NICHD SECC website, 2011). Interviewers visited mothers and

children in their homes when children were 1, 6, 15, 24, 36, and 54 months old. Families also

participated in telephone interviews between home visits that were conducted at 3, 9, 12, 18, 21,

27, 30, 33, 42, 46 and 50 months. During the home visits, mothers completed a variety of

instruments designed to measure depression, parenting stress, relationship quality, social support,

and their attitudes towards work and child-rearing. Mothers also provided extensive information

about socio-demographic characteristics, employment, and child care during the home visits.

During the home visits, trained interviewers observed mothers’ interactions with children and the

state of the household environment. Information on household composition, employment, child

care and some health measures were updated every 3 to 4 months through the telephone contacts.

Beginning with the 15-month interview, certain mother/child assessments were conducted in a

laboratory setting.

In our first set of analyses, we analyze effects of maternal work on family outcomes

measured when children are 6 months old. Our 6-month sample includes 1,198 mother/child

pairs who have available information on all measures used in the analysis, with the exception of

maternal occupation prior to childbirth, maternal reading score, and maternal smoking during

pregnancy. For these latter three measures (described below), we replaced missing values with

sample means and included in all models a dummy variable indicating that an imputed value was

used.

11

In our second set of analyses, we use a pooled sample which includes repeated

observations on children and mothers. When analyzing effects of maternal employment on

maternal overall health, we pool data from all telephone and home interviews from 1 month until

54 months, yielding potentially 16 assessment time points for each mother (we are not able to

use the 1 month assessment as a time point since we use this interview for lagged values of the

dependent variable for the 3 month interview, as described below). In these models, maternal

overall health and maternal employment are measured about every 3 months. We run analyses

with an unbalanced panel of 18,655 observations which includes observations with available data

on maternal health, employment, a standard set of time-invariant characteristics.

When analyzing maternal depression and parenting quality, we pool data from the 6, 15,

24, 36 and 54 month interviews. For the depression outcomes, we run analyses on a sample

limited to 5,618 observations with available data; for parenting quality, the sample size is 4,371.

When analyzing parenting stress, we pool data from the 6, 15, 24 and 36 month interviews, since

parenting stress is not assessed at the 54 month interview – the sample size for these analyses is

4,573. Results are discussed below.

a. Family Outcomes

Depressive symptoms: We measure maternal depression using the 20-item Center for

Epidemiologic Studies Depression Scale (CES-D), which is used to measure depressive

symptoms in the past week in non-clinical populations. Mothers completed CES-D instruments

during the 1, 6, 15, 24, 36, and 54 month home interviews. The CES-D is one of the most widely

used psychiatric scales and captures mood, somatic problems, problems in interactions with

others, and issues with motor functioning, such as “I felt lonely,” “my sleep was restless,” and “I

could not get going.” The respondent is asked to respond to each item according to a 4-point

12

Likert scale, with higher values corresponding to higher frequency of the item in the past week.

For example, for the item “I felt lonely,” mothers responded either “less than 1 day” (zero

points), “1-2 days” (1 point), 3-4 days (2 points), or 5-7 days (3 points). Scores range from 0 to

60, and a score of 16 or higher is suggestive of clinically defined depressive disorder. The CES-

D scale, however, does not correspond to a DSM-IV diagnosis of major depression. It is used

primarily as a screening tool for depression, not as a diagnostic tool (Eaton et al. 2003).

We create two measures of depression from the CES-D scale, a continuous measure of

symptoms and a dichotomous indicator of depression. Because the CES-D is skewed to the right

in these data, we use the natural log of the total CES-D score as the continuous measure.2 The

dichotomous measure is a dummy variable indicating whether or not the respondent’s CES-D

score is equal to or exceeds 16. This dummy variable is not equivalent to a psychiatric diagnosis

of depression, but it does capture respondents who are experiencing many symptoms of

depression, or several symptoms with high frequency, in the past week (Eaton et al. 2003).

The sample average CES-D score at the 6-month interview is 8.9, and 17 percent of the

sample is depressed at the time of the 6-month interview (Table 1). It is notable that mothers

who are employed at 6 months have appreciably lower CES-D scores and rates of depression at

both the 1 month and 6 month interviews compared to mothers who are not employed at 6

months (Table 2). The rate of depression among full-time employed mothers at 6 months is 15

percent versus 22 percent among mothers who were not employed at 6 months (Table 2).

Overall health: Every three months, SECC mothers rated their own health in the past 3

months, compared to other women their age. Mothers can report their health as poor (1), fair (2),

good (3) or excellent (4). We combine the poor and fair rankings since the number of mothers

reporting poor health was small. We use this rating as an outcome measure, as well as a 2 In this variable and in others where log values are used, the zeros are replaced with a value of 0.5.

13

dichotomous indicator that equals one if the mother reports her health in general is fair or poor.

Since the question does not specify physical or emotional health, these variables may capture

both physical and mental illness.

Mothers reported on their health through home and telephone interviews every 3-4

months from the 1 month until the 54 month interview. At the 6-month interview, 12 percent of

mothers report poor or fair health (Table 1). Employed mothers at 6 months have much lower

rates of poor/fair health at the 1 month and at the 6 month interviews compared to mothers who

are not employed at 6 months – for example, 9 percent of full-time employed mothers report

poor or fair health at 6 months compared to 18 percent of mothers who are not employed at 6

months (Table 2).

Parenting Stress: To measure parenting stress, we draw on two scales completed by

SECC mothers during the home interviews. At the 1 month and 6 month interviews, mothers

completed a 30-item version of the Abidin Parenting Stress Index, which is designed to measure

parent-child relationship stress and risk for adverse parenting and child behavioral outcomes.

The index includes items such as “I feel trapped by my responsibilities as a parent”, “I enjoy

being a parent” and “I feel capable and on top of things when caring for my baby.” At the 15, 24

and 36-month interviews, mothers were administered a 20 item adapted form of the Parent Role

Quality Scale, which is appropriate to measure parenting stress among parents of toddlers and

pre-school age children. Mothers are presented with ten potential concerns and ten potential

rewards of child-rearing, and are asked to rate how much these concerns and rewards reflect their

own experiences in parenting. The scale includes concerns such as “feeling tied down because

of the children” and “the unending responsibilities” and rewards such as “the love your child

shows” and “seeing your child grow and change.” For both measures of parenting stress, higher

14

scores indicate a greater degree of parenting stress. Table 2 shows that employed mothers at 6

months report lower rates of parenting stress than mothers who were not employed at 6 months.

This is true of parenting stress measured at the 1 month interview as well.

Maternal Sensitivity: Maternal sensitivity is measured using trained observers’ ratings of

videotapes of mothers’ behavior toward their children in semi-structured play situations. At 6

and 15 months, mother/child interactions were observed in the child’s home, while at 24, 36, and

54 months these observations were conducted in the laboratory. These interactions are designed

to demonstrate the degree to which the mother responds in a sensitive way to the child’s

nondistress, intrusiveness (reverse scored), and positive regard (at the 6, 15, and 25 month

assessments), and the mother’s supportive presence, hostility (reverse scored) and respect for

autonomy (at 36 and 54 months). Higher scores indicate higher degree of sensitivity to the child.

Table 2 shows that although employed mothers report better mental and overall health and lower

parenting stress than mothers who are not employed at 6 months, full-time employed mothers are

not different from mothers who are not working at 6 months in maternal sensitivity. Part-time

working mothers have slightly higher sensitivity ratings compared to mothers who are not

working at 6 months.

b. Maternal employment

Mothers provided employment information during each home interview (1, 6, 15, 24, 36

and 54 months) and most this information was updated during intervening telephone contacts

every 3 to 4 months. For each of these potentially 16 time points, we created several measures of

lagged maternal employment. As described in the next section, we use lagged measures to avoid

problems with reverse causality. The measures are: (1) the number of hours worked per week

measured at the home or telephone interview that was conducted 3 months prior (hours worked

15

at last interview); (2) the average of weekly hours prior two assessments conducted 3 and 6

months ago (average hours worked over past 2 waves); and (3) the average weekly hours the

mother worked up to and including the most recent prior assessment point (average hours

worked in child’s life). Based on these continuous measures of work hours, we also created

three dummy indicators for: (1) worked 1 to 20 hours per week at last interview; (2) worked 21

to 39 hours per week at last interview; and (3) worked at least 40 hours per week at last

interview. When we estimate models using the full sample, the baseline category combines

mothers who are not employed with mothers who are employed but are on leave at 3 months.

When we estimate models using a sample limited to employed mothers (defined as mothers who

report that they are employed and working or employed and on leave at the 1 month interview),

the baseline category is limited to mothers who are employed but are still on leave at the 3 month

interview.

In addition to work hours, employed SECC mothers provided information regarding the

characteristics of their jobs. In all home and telephone interviews from 3 to 54 months, mothers

were asked the number of hours they could work at home, and the time of day of their work

hours (daytime, evening, night or varying shifts). From this information, we created a dummy

variable indicating the mother could work at least 10 hours a week from home, and a dummy

variable indicating the mother worked a non-daytime shift. During the 6, 15, 24 and 36 month

interviews, mothers also reported whether their job required overnight travel (never, less than

once a month, more than once a month) and whether their work hours were not at all flexible, a

little flexible (can leave in an emergency), fairly flexible, and completely flexible. From this

information, we created dummy variables indicating whether the mother’s job requires any

overnight travel (either less than once a month or more than once a month) and whether the

16

mother’s job is either fairly or completely flexible. When work characteristics are of interest, we

limit the sample to employed mothers.

At the 6-month interview, 64 percent of mothers were employed or on leave, and the

average weekly hours among employed mothers was about 33 (Table 1). Among employed

mothers at the 6-month interview, 64 percent reported flexible hours, 22 percent worked

alternate shifts, 18 percent had overnight travel, and 12 percent worked from home at least 10

hours per week (Table 1).

c. Other covariates

To adjust for other factors that may confound an association between maternal

employment and family outcomes, we estimate models that include extensive sets of controls for

family socio-economic status, the mother’s education and ability, the mother’s attitudes towards

employment, and the initial health endowment of the child. All models include the following

measures: mother’s age in years, number of years of education, size of household measured at 1-

month interview, maternal race/ethnicity (dummy indicators for African-American and Other

race with white as the baseline, dummy indicator for Hispanic), child’s gender (dummy indicator

for female), dummy indicator for child’s birth order (second, third, fourth, or higher with

firstborn as the baseline); dummy indicators of birth month of the child (all were born in 2001);

dummy indicator for low birth-weight child (2500 grams or less); dummy indicator for

premature child (born before 37 weeks gestation); dummy indicator for whether mother smoked

at all during pregnancy; dummy indicator for any pregnancy complications; mother’s

standardized score on PPVT reading test administered at 36 month interview, and dummy

indicators for each SECC site.

17

Some models also include the following measures: dummy indicators for mother’s

occupation in the year prior to childbirth,3 family income in the year prior to childbirth, current

family income, mother’s score on scale measuring progressivity of child-rearing beliefs

measured at 1-month interview, mother’s score on scale measuring work commitment

administered at 1-month interview, mother’s score on scale measuring benefits of maternal

employment administered at 1-month interview, dummy indicators for family structure at 1-

month (single parent/other family structure, live-in partner with married as the baseline, dummy

indicator for child’s father does not live in HH), and dummy indicators for whether the child

spends at least 10 hours a week in six types of child care arrangements (daycare center, child

care home, father, grandparent, in-home caregiver, multiple arrangements).4

IV. Empirical Approach A. Effects of maternal employment on 6-month family outcomes

We begin by estimating Eq. 1 using outcomes measured when children are 6 months old

(t=6 in Equation 1).

(1) yi6 = Xi6α + Wi3β + νi + εi6

In equation 1, yi6 represents the ith family’s health or parenting outcome at the 6-month

interview, X is a vector of covariates, W is a vector of lagged measures of maternal employment

measured at the 3-month interview, ν represents unmeasured family/maternal characteristics that

affect y, ε is a random error term, and α and β are parameters to be estimated. As discussed

below, lagged employment is used to help reduce the possibility of bias resulting from reverse

3 Professional; technician or related support; sales; administrative support or clerical; private household; protective service; service; farm operation or management; mechanic or repairer; construction or other trade; machine operator, assembler, or inspector; transportation or material moving; handler, equipment cleaner, helper, or laborer; with executive, administrative or managerial as the baseline. 4 Children who did not use any of these child care arrangements for 10 or more hours per week were considered to be exclusively in the care of the mother. At 6 months, about 14% of employed mothers were using exclusive maternal care.

18

causality. For continuous outcomes (e.g., depressive symptoms, parenting quality, parenting

stress), we use OLS for estimation of Equation 1, while we use standard probit models for binary

outcomes (e.g. poor/fair health, depressed) and an ordered probit model for self-assessed health

rating (e.g., 1 = fair/poor health, 2 = good health, 3 = excellent health). We estimate Huber-

White standard errors adjusted for clustering on site.

The causal effect of maternal employment on family wellbeing, β, may be positive or

negative. Maternal employment may detract from family wellbeing by reducing the amount of

time mothers spend investing in their families and in their own health and wellbeing. Prior

research indicates that employed mothers spend less time with their children than non-employed

mothers (Cawley & Liu, 2007).5 On the other hand, maternal employment brings more income

that can be used to purchase market goods that benefit the family and the mothers themselves.

Estimation of β is complicated by several potential problems. First, if families/mothers

have unmeasured characteristics (ν) that are correlated with both family outcomes (y) and

maternal work hours (W), the estimate of β will be biased. As Ruhm (2004), Gregg et al. (2005),

and others have noted, this bias could operate in either direction. Mothers who are relatively

resistant to stress and depression may be more likely to work, and also more likely to have

positive family outcomes, compared to less resistant mothers. This case would suggest that β is

biased upwards, since the coefficient captures both the causal effect of employment on outcomes

as well as the unmeasured resilience of mothers.

On the other hand, if mothers select into employment along unmeasured characteristics

that detract from family outcomes, the estimate of β would be biased downwards. For example,

5 Cawley & Liu (2007), using 2003-2006 data from the American Time Use Survey, find that conditional on spending some time with children, employed mothers spend 139 fewer minutes (on the reference day) with their children than stay-at-home mothers, controlling for maternal race/ethnicity, age, education, marital status and other factors (Cawley & Liu, 2007). Bianchi (2000) notes that employed mothers sleep fewer hours and spend less time in self-care and leisure time activities compared to non-employed mothers.

19

if mothers experiencing difficult adjustment to parenthood are more likely to return to work and

more likely to have adverse outcomes, β will capture both the effects of employment and this

poor adjustment effect. In our sample, employed mothers at 6 months tend to be healthier at

baseline and generally more advantaged than non-employed mothers (Table 2). This fact

suggests, but does not definitively imply, that employed mothers also may have unobserved

characteristics that pre-dispose them to favorable family outcomes compared to non-employed

mothers (Altonji et al., 2005).

In order to reduce the possibility of bias from unmeasured factors, we estimate models

with unusually rich and extensive sets of control variables. We begin with a specification that

includes only pre-determined variables – maternal age, maternal race/ethnicity, total household

size, maternal education, maternal reading test score (a proxy for ability), pregnancy

complications, mother smoked during pregnancy, low birth weight child, premature child, child

gender, birth order of child, birth month of child, site fixed effects, and lagged maternal health

and parenting (measured by maternal depression, maternal overall health and maternal parenting

stress at the 1-month interview). Notably, our baseline model adjusts extensively for the

family’s initial state of health and wellbeing, as well as for maternal ability, which are perhaps

the two most likely confounding factors.6

Next, we estimate richer specifications to gauge the degree of selection along observed

characteristics of the family. We incrementally add controls for (1) maternal occupation prior to

the child’s birth and family income prior to the child’s birth; and (2) family structure, maternal

beliefs about the benefits of employment, work commitment, and beliefs about child-rearing

6 Although it is possible that maternal employment already has affected family outcomes by the 1 month interview, this is unlikely. In our sample, 135 mothers are employed at the 1-month interview, but only 31 of these mothers have returned to work full-time, and 104 of these mothers report having fairly or completely flexible hours.

20

measured when the child was 1 month old.7 We also re-estimate all of these models after

limiting the sample to mothers who are employed (either working or on leave) at the 1 month

interview. Limiting the sample this way may further reduce unobserved heterogeneity, although

it comes at the cost of a reduction in power.

In addition to unobserved heterogeneity, a second potential problem that arises in

estimating β is reverse causality. Although we seek to estimate the effect of maternal

employment on family wellbeing, causation may run the other way, with mothers changing their

employment decisions in response to their own and their family’s health and wellbeing. For

example, a mother may cut back on her work hours if she feels that employment interferes with

her parenting or causes stress for the family. To address the possibility of reverse causality, we

use a lagged measure, maternal employment at 3 month interview, in all 6-month models.

A third empirical issue is the interpretation of the estimate of β. Our interest is in the

total effect of maternal employment on family outcomes (Gregg et al., 2005). For example, if

maternal employment leads to usage of child care and an increase in family income, these

changes may affect families, and our intention is to capture both types of effects through the

coefficient on maternal employment. This focus affects what right hand side variables we

include in the model. Although we include an extensive set of controls to reduce unobserved

heterogeneity, we avoid including variables that may capture mechanisms through which

maternal employment affects families.

In particular, in our main specifications, we do not include current family income or

current child care arrangements as right hand side variables since these factors may be critical

mechanisms through which maternal employment affects family outcomes. In alternative

7 These latter two sets of variables, particularly those measured when the child is 1 month old, may be endogeneous to the return-to-work decision. Thus, we interpret findings from these two models with caution.

21

specifications, however, we do include these variables in order to gauge how much of the effect

of employment on family outcomes is operating through income and child care arrangements.

Income is likely to be the primary route (although probably is not the only route) through which

maternal employment improves family outcomes; including a control for family income thus

may exacerbate any negative effects of maternal employment that may exist. In the case of child

care, one possibility is that the use of child care improves family outcomes, if child care is a

positive experience for children and their parents – in this case, including controls for child care

would, like income, worsen any negative effects of maternal employment that may exist. If child

care causes stress and difficulties for children and parents, however, including controls for child

care may actually reduce any estimated negative effects of maternal employment on outcomes.

A fourth empirical issue is the possibility that the effect of maternal employment varies

by family characteristics. Following the literature on maternal employment and children, we

estimate models on the following sub-samples: mothers who are living in poverty at baseline vs.

mothers who are not; mothers with less than 16 years of education vs. college educated mothers;

mothers living in poverty at baseline vs. mothers not living in poverty at baseline; and married

vs. unmarried mothers. Since the SECC did not over-sample disadvantaged or minority

populations, our samples sizes are small for some of these analyses, and we cannot examine

effects for racial/ethnic sub-samples.

B. Effects of maternal employment on family outcomes during the first 4.5 years of life

Next, we move to examining whether the effects of maternal employment on family

outcomes persist during early childhood. To do so, we use pooled data from Phases I and II of

the SECC, which include information collected when the children were 1 month to 54 months

22

old. Here, we examine effects of lagged maternal work hours only, and do not consider other

work characteristics, due to data availability.

We estimate dynamic panel data models which account for individual fixed effects, as

well as the strong degree of persistence in maternal health and parenting outcomes during the

first 4.5 years of children’s lives. This persistence may be due to true state dependence (e.g.,

stock of maternal health and parenting capital evolves slowly over time) and/or to unmeasured

factors (e.g. genetics) (Contoyannis, 2004). To account for this persistence, our models include a

lagged dependent variable on the right hand side, as seen below in Equation (1’). In equation 1’,

yit represents the ith family’s health/parenting outcomes at time t, yit-1 is a lagged dependent

variable measured 3 months prior to t, X is a vector of time-varying covariates, W is a lagged

measure of maternal employment measured 3 months prior to t, ν represents unmeasured

family/maternal characteristics that affect y, ε is a random error term, and α, β, and δ are

parameters to be estimated.

(1’) yit = yit-1δ + Xitα + Wit-1β + νi + εit

In the case of the maternal overall health outcome, we draw on unusually rich data.

Maternal overall health, maternal employment information and some critical time-varying

factors (household size, whether the father lives in the household) are reported by mothers about

every 3 months from the 1 month to the 54 month interviews. We emphasize that the frequency

of the data collection for this outcome reduces the likelihood of confounding by unmeasured

time-varying factors, but increases the likely degree of state dependence in the maternal health

outcome. For other outcomes (depression, parenting stress, quality of parenting), we draw on

data that is available at the 1, 15, 24, 36, and 54 month interviews. Confounding by time-varying

unmeasured factors is of greater concern in these models.

23

To estimate Eq. 1’, we use Arellano-Bond (A-B) difference GMM methods (Arellano &

Bond, 1991). These methods are well-suited for our case in which we have: (1) individual fixed

effects; (2) a lagged dependent variable on the right hand side; and (3) a large N and a small t.

A-B methods address the problem that lagged health/parenting (on the right hand side of Eq. 1’)

is correlated with the error term once the equation is first-differenced to remove the individual

fixed effect. The A-B approach involves instrumenting for the difference between lagged and

current health/parenting using deeper lagged measures of health/parenting. In our application,

we also need to instrument for lagged work hours, which are predetermined in Equation 1’ but

become endogenous to the model when we take the first difference. We use the one-step version

of the A-B estimator which generates robust standard errors, and we report results from tests for

first and second order autocorrelation, and from Hansen’s J test of overidentification.

V. Results

We begin the formal analysis with an examination of the effects of maternal work hours

measured at the 3 month interview on the five different maternal mental health and parenting

outcomes measured when the child is six months old. Table 3a shows results for the full sample,

where mothers who are employed but are on leave at 3 months have zero work hours and thus

are combined in the baseline category with mothers who are not employed. Table 3b shows

results for the employed sample, where mothers with zero work hours at 3 months are all

employed but on leave from their jobs. In both tables, the dependent variables are listed in the

rows. Column 1 shows estimates from the most parsimonious specification. The models in the

columns labeled 2 through 6 incrementally add covariates as indicated in the bottom rows of the

table. For continuous outcomes, we show estimated coefficients and T-statistics from OLS

24

models, and for dichotomous outcomes, we show average marginal effects and T-statistics from

probit models.

The first row of estimates in Tables 3a and 3b pertain to the log CES-D score. For both

the full and the employed samples, lagged hours of work are associated with increases in

depressive symptoms. While the magnitude varies some according to the other included

covariates, all coefficients are positive and statistically significant at the 5 percent level in a two-

tailed test. The coefficients indicate that on average, an increase in weekly work hours of 10

hours is associated with an increase in the depression score in the range of 3 to 7 percent among

the full sample, and a similar range of 6 to 9 percent for the employed sample. These are

relatively small effects, given that a 10 hour increase corresponds to about a 40 percent increase

in work hours among employed mothers (lagged work hours among employed mothers at the 6

month interview was about 25 hours per week).

Moreover, increases in work hours do not move mothers over the indicator threshold for

depression unless current child care arrangements are included as covariates and the full sample

is used for estimation (Tables 3a and 3b). Columns 5 and 6 in Table 3a show that after the

indirect effect of child care arrangements is netted out of the total effect, increased work hours

have a detrimental effect on maternal mental health, as indicated by an increased probability of

being over the CES-D threshold score of 16. The magnitude of this effect implies that a ten hour

increase in work hours on average increases the likelihood of depression by .02 in the both

samples, which is about a 12 percent increase at the sample mean for depression of 17% in the

full sample (Columns 5 and 6, Table 3a).

But what this finding also indicates is that that at least some forms of child care have a

protective effect on maternal mental health – adjusting for other factors, the results show that

25

using center-based child care and having a grandparent care for the child, relative to care by the

mother, both are associated with reductions in depressive symptoms at the 6 month interview

(results not shown). However, as we argue above, the effect we are most interested in is the total

effect of work hours on the outcomes, not the effect of working net of the mechanisms. The

models that exclude child care represent the total effect of hours on the depression indicator,

which includes the direct effect of working on health, as well as the indirect effects of child care

arrangements and income (which is determined by work hours) on health.

For the full sample, Table 3a shows that overall health, parenting stress, and parenting

quality are all unaffected (statistically) by increases in maternal work hours. These insignificant

results persist regardless of the other included variables. However, these results likely reflect the

averages of two groups of women, those who work and those who do not. Among employed

mothers, in fact, we find that hours worked is positively associated with self-reported parenting

stress at 6 months (Table 3b). However, this stress apparently does not translate into poorer

parenting based on objective measurements, as increased work hours are not statistically

associated with the parenting quality (maternal sensitivity) score at 6 months (Tables 3a-b).

We also estimated the models shown in Tables 3a-b using sub-samples of mothers who

were college educated, not college educated, married at the 1 mo. interview, not married at the 1

mo. interview, living in poverty at the 1 mo. interview and not living in poverty at the 1 mo.

interview (results not shown). Some of these samples are small, but the general pattern of

findings suggests that the negative effects of work hours on depression and parenting stress are

driven by effects among married mothers, and they hold for both college and not college

educated mothers. The sub-sample analyses indicate that after including controls for income and

26

child care, work hours detract from parenting quality in the not college educated and the not poor

samples, but not in the college educated or the poor samples.

In Table 4, we show results from ordered probit models in which we examine the effects

of maternal work hours on self-rated overall health of mothers measured at 6 months. The

findings suggest that in the full sample, which includes both working and non-working mothers,

there are no effects of work hours at 3 months on maternal overall health at 6 months. However,

among mothers who are employed or on leave at 6 months, higher work hours at 3 months is

associated with a reduction in overall health. This effect is small in magnitude – a 10 hour

increase in work hours at 3 months (about a 40 percent increase) is associated with a 2 percent

reduction in the probability of being in excellent health, and a 2 percent increase in the

probability of being in good health.

In the appendix, we show results from models in which we use alternative definitions of

work hours, and results from models in which we examine characteristics of work at 6 months.

Rather than use the number of weekly hours worked at the 3 month interview, in Appendix Table

1 we use 1) the average hours per week worked over the child’s life (i.e. in the first 6 months); 2)

the average weekly hours reported in the one month and three month interviews; and 3) three

indicator variables for the categories of working 1 to 20 hours per week at last interview,

working 21 to 39 hours per week at last interview, and working at least 40 hours per week at last

interview. The models shown in these tables use the same set of covariates used in column 3 of

Tables 3a and 3b. We use this specification because it includes the most complete set of

covariates, while still providing an estimate of the total effect of work hours on the outcomes.

The results in Appendix Table 1 are similar to those in the previous tables, where more

hours of work are associated with higher CES-D scores. However, we learn from this table that

27

the effect is driven by, and is concentrated among, women working 40 or more hours at the three

month interview. Mothers who work 40+ hours when their infants are 3 months old have

depressive symptoms score that are 16 to 22 percent higher than other mothers, after adjusting

for other factors (Appendix Table 1). This increase in depressive symptoms, however, does not

translate into an increase in the likelihood of having a CES-D score greater than 16 (Appendix

Table 1). The probability of being in poor health is not associated with work hours, nor is the

parenting quality score. Parenting stress, however, is increasing with hours of work among

women in the employed sample. Women who work any of the categories of hours (1-20, 21-39

and 40+) at 3 months report more stress relative to their employed counterparts who are on leave

and working zero hours at 3 months (Panel B, Appendix Table 1). This result does not hold in

the corresponding model using the full sample (Panel A, Appendix Table 1).

Appendix Table 2 shows the effects of different job characteristics on the maternal health

and parenting outcomes among the sample of employed women. The job characteristics

considered are: whether flexible hours are available, whether the job requires overnight travel,

whether the mother works non-standard hours, and whether the mother works from home 10

hours or more a week. We interpret these findings with caution since these characteristics are

measured at 6 months (the mothers did not provide information on flexible hours and overnight

travel at the 3 month telephone interview). Two findings are notable. First, more hours of work

at 3 months are still associated with more depressive symptoms and parenting stress at 6 months,

even when controls for work characteristics are included in the model. Second, although flexible

hours and overnight travel are not statistically associated with any of the health and parenting

outcomes, non-standard work hours are associated with increases in depressive symptoms of 16

to 18 percent, as well as increases in the parenting stress index. The ability to work from home

28

at least 10 hours per week is associated with a reduction of in the probability of being in fair or

poor health, although the size of this effect appears to be implausibly large.

To summarize, the results thus far demonstrate that work hours, particularly full-time

work hours, at 3 months have adverse effects on maternal depressive symptoms, parenting stress,

and overall maternal health measured at 6 months. But these effects do not appear to have

clinical ramifications since work hours are not associated with the quality of parenting (as

measured by a trained observer) or dichotomous indicators of depression and fair/poor health.

The next question to ask is if these detrimental effects persist when we examine the

child’s first 4.5 years of life (1 mo. – 54 mo. interview). In Table 5, we present two sets of

findings estimated using a sample that pools data from the 1 month to the 54 month interviews.

The first set of findings are estimates from standard OLS models with robust standard errors

adjusted for clustering on respondent id. These models are based on our preferred specification

(column 3 in Tables 3a-b) but they also include a lagged version of the dependent variable as a

covariate. The obvious drawback of these models is omitted variables bias - the individual fixed

effect may be directly related to the health/parenting outcome, as well as correlated with right

hand side variables, including lagged work hours. The next set of findings show the A-B

estimates. In these models, we account for both state dependence as well as unobserved

heterogeneity, and we treat the dependent variables as continuous in all cases.

In Table 5, two findings are notable. First, for all outcomes, the pooled OLS models

show strong state dependence. However, once the individual fixed effect is acknowledged in the

A-B model, state dependence is no longer apparent in the log CES-D, parenting stress, and

parenting quality outcomes. In the case of the depressed, poor/fair health, and overall health

outcomes, state dependence still exists, but its importance is diminished considerably once fixed

29

effects are considered. Second, there is no evidence that the detrimental effects of maternal

employment we observed at the 6 month interview persist during early childhood. In fact, more

maternal work hours is associated with reductions in parenting stress, perhaps because mothers

find combining work and child care satisfying once children grow older. More maternal work

hours is associated with lower overall health, but the magnitude of this effect appears to be small,

and we find no effect of work hours on the likelihood of being in fair/poor health.

As expected, we reject the null of zero autocorrelation in the test for order 1

autocorrelation. We fail to reject order 2 autocorrelation in all models except the parenting stress

model. This result suggests no autocorrelation in the original error term in any of the models

aside from the parenting stress model. We fail to reject the overidentifying assumptions in all

models expect parenting stress and log CES-D. Overall, then, the A-B model appears

appropriate in all models expect for parenting stress, and we have some concern about the

identifying assumptions in the CES-D model as well.

VI. Conclusions

Prior work on the effects of maternal employment has focused on child outcomes. In this

paper, we add to the literature by examining effects of maternal work on maternal health and

parenting outcomes. We find that among employed mothers, work hours measured when infants

are about 3 months old are positively associated with depressive symptoms and parenting stress,

as well as a small decline in self-reported overall health, measured when infants are about 6

months old. The effects on depression are driven by mothers who are working full-time at 3

months post-childbirth, while for parenting stress, any level of work hours at 3 months is

associated with adverse effects. The only work characteristic that appeared to affect outcomes

30

was alternate shift work, which is associated with more depressive symptoms and parenting

stress at 6 months.

However, these detrimental effects do not persist as children get older. During the first

4.5 years of children’s lives as a whole, more work hours is not associated with depression, and

has very small detrimental effects on maternal overall health. If anything, more maternal work

hours reduces parenting stress during the first 4.5 years.

These findings are consistent with those of Brooks-Gunn et al. 2010, who use structural

equation modeling and data from the SECC to assess the effects of first-year maternal

employment on children’s outcomes up to age 7. They find that there are both negative and

positive effects of maternal employment on children’s cognitive, social, and emotional outcomes,

but the net effect on children is zero (Brooks-Gunn et al., 2010). We find that for maternal

health and parenting outcomes, full-time maternal employment at 3 months detracts from

mothers’ health and increases stress at 6 months post childbirth. These effects are unlikely to

have long-term effects on families since (1) maternal employment affects depressive symptoms,

but does not affect the CES-D threshold for a likely clinical case of depression; (2) the effects on

maternal overall health are statistically significant, but small in magnitude; and (3) self-reported

parenting stress is affected, but objective measurement of parenting quality is not affected.

Moreover, when the first 4.5 years post-childbirth are considered as a whole, there is no pattern

of negative effects of maternal employment on maternal health and parenting.

However, the results still suggest that the transition back into employment immediately

after childbirth is difficult for the average family, detracting from maternal health and increasing

self-reported parenting stress. These findings emphasize the need for parental leave policies that

31

allow new parents to take longer leave, and/or work fewer hours in the first few months after

childbirth.

32

References 1. Altonji, J.G., Elder, T.E., Taber, C.R., 2005. Selection on Observed and Unobserved

Variables: Assessing the Effectiveness of Catholic Schools. Journal of Political Economy 113, 151-184.

2. Arellano, M., Bond, S., 1991. Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies 58(2), 277-297.

3. Baker, M., Milligan, K., 2008. Maternal employment, breastfeeding, and health: Evidence from maternity leave mandates. Journal of Health Economics 27, 871-887.

4. Baker, M., Gruber, J., Milligan, K., 2008. Universal Child Care, Maternal Labor Supply, and Family Well-Being. Journal of Political Economy 116, 709-856.

5. Barry TM, Dunlap ST, Cotton SJ, Lochman JE, Wells KC. 2005. The influence of maternal stress and distress on disruptive behaviors problems in boys. Journal of the Academy of Child and Adolescent Psychiatry 44: 265-273.

6. Belsky J. 1988. The ‘effects’ of infant day care reconsidered. Early Childhood Research Quarterly 3: 235-272.

7. Berger, L.M., Hill, J., Waldfogel, J., 2005. Maternity Leave, Early Maternal Employment and Child Health and Development in the US. The Economic Journal 115, F29-F47.

8. Berger, L., Brooks-Gunn, J., Paxson, C., Waldfogel, J., 2008. First-year maternal employment and child outcomes: Differences across racial and ethnic groups. Children and Youth Services Review 30, 365-387.

9. Baum, C.L. (2003). Does early maternal employment harm child development? An analysis of the potential benefits of leave taking. Journal of Labor Economics 21: 409-48.

10. Bianchi, S.M., 2000. Maternal Employment and Time with Children: Dramatic Change or Surprising Continuity? Demography 37, 401-414.

11. Bornstein, M.H. (ed.), 2002. Handbook of Parenting, Biology and Ecology of Parenting. Lawrence Erlbaum Associates. Place of Publication: Mahwah, NJ.

12. Brooks-Gunn J, Han WJ, Waldfogel J. 2002. Maternal employment and child cognitive outcomes in the first three years of life: the NICHD Study of Early Child Care. Child Development 73 : 1052-72.

13. Brooks-Gunn, J., Markman, L.B., 2005. The Contribution of Parenting to Ethnic and Racial Gaps in School Readiness. Future of Children 15, 139-168.

14. Brooks-Gunn, J., Han, W.J., Waldfogel, J., 2010. First year maternal employment and child development in the first 7 years. Monographs of the Society for Research in Child Development, 75(2), pp. 1-147.

15. Cameron, A.C., Trivedi, P.K., 2005. Microeconometrics Methods and Applications Cambridge University Press, New York.

16. Campbell SB, Cohn JF, Meyers T. 1995. Depression in first-time mothers: Mother–infant interaction and depression chronicity. Developmental Psychology 60 : 349-357.

17. Cawley, J., Liu, F., 2007. Maternal Employment and Childhood Obesity: A Search for Mechanisms in Time Use Data. NBER Working Paper 13600.

18. Chatterji, P., Markowitz, S., 2005. Does the length of maternity leave affect maternal

33

health? Southern Economic Journal 72, 16-41.

19. Chatterji, P., Markowitz, S., 2008. Family leave after childbirth and the health of new mothers. NBER Working Paper 14156.

20. Cohn J, Matias R, Tronick EZ, Connell D, Lyons-Ruth K. 1986. Face-to-face interactions of depressed mothers and their infants. New Directions for Child Development 34: 31-45.

21. Coleman, J., 1988. Social capital in the creation of human capital. American Journal of Sociology, 94, S95 – S120.

22. Collins, W. A., Maccoby, E. E., Steinberg, L., Hetherington, E. M., & Bornstein, M. H. (2000). Contemporary research on parenting: The case for nature and nurture. American Psychologist, 55, 218–232.

23. Contoyannis, P., Jones, A., Rice, N., 2004. The dynamics of health in the British Household Panel Survey. Journal of Applied Econometrics, 19: 473-503.

24. Dearing, E., Taylor, B.A., McCartney, K., 2004. Implications of Family Income Dynamics for Women’s Depressive Symptoms During the First 3 Years After Childbirth. American Journal of Public Health 94, 1372-1377.

25. Daniel, S.S., Grzwacz, J.G., Leerkes, E., Tucker, J., Han, W.. Nonstandard maternal work schedules during infancy: Implications for children’s early behavior problems. Infant Behavior & Development 32(2), 195-207.

26. Downey, G., Coyne, J.C., 1990. Children of depressed parents: An investigative review. Psychological Bulletin 108, 50-76.

27. Eaton WW, Muntaner C, Smith C, Tien A, Ybarra M. 2003. Center for epidemiologic studies depression scale: Review and revision (CESD and CESDR). In The use of psychological testing for treatment planning and outcomes assessment, 3rd edition, chapter 40, volume III, edited by M.E. Maruish. Mahwah, NJ : Lawrence Erlbaum Associates, pp.8-9.

28. Essex M, Klein M, Miech R, Smider N. 2001. Timing of initial exposure to maternal major depression and children’s mental health symptoms in kindergarten. British Journal of Psychiatry 179: 151-156.

29. Feldman R, Sussman AL, Zigler E. 2004. Parental leave and work adaptation at the transition tp parenthood: Individual, marital, and social correlates. Journal of Applied Developmental Psychology 25(4) : 459-479.

30. Field, T. 1984. Early interactions between infants and their postpartum depressed mothers. Infant behavior and Development 7: 517-522.

31. Gray RF, Indurkhya A, McCormick MC. 2004. Prevalence, stability, and predictors of clinically significant behaviors problems in low birth weight children at 3, 5, and 8 years of age. Pediatrics 114 : 736-743.

32. Gregg, P., Washbrook, E., Propper, C., Burgess S., 2005. The Effects of a Mother’s Return to Work Decision on Child Development in the UK. The Economic Journal 115, F48-F80.

33. Han, W., 2005. Maternal nonstandard work schedules and child cognitive outcomes. Child Development 76, 137-154.

34

34. Hay DF, Pawlby S, Angold A, Harold GT, Sharp D. 2003. Pathways to violence in the children of mothers who were depressed postpartum. Developmental Psychology 39 : 1083-1094.

35. Hill, J.L., Waldfogel, J., Brooks-Gunn, J., Han, W., 2005. Maternal Employment and Child Development: A Fresh Look Using Newer Methods. Developmental Psychology 41, 833-850.

36. Ispa JM, Fine MA, Halgunseth LC, Harper S, Robinson J, Boyce L, et al. Maternal intrusiveness, maternal warmth, and mother-toddler relationship outcomes: Variations across low-income ethnic and acculturation groups. Child Development. 2004;75:1613–1631.

37. James-Burdumy, S., 2005. The Effect of Maternal Labor Force Participation on Child Development. Journal of Labor Economics 23, 177-211.

38. Kim-Cohen, J., Caspi, A., Moffitt, T.E., 2005. Prior juvenile diagnoses in adults with mental disorder. Archives of General Psychiatry 60, 709-717.

39. Lyons-Ruth, K., 1996. Attachment relationships among children with aggressive behavior problems: the role of disorganized early attachment patterns. J Consult Clin Psychol 64, 64-73.

40. Minkovitz, C.S., Stobino, D., Scharfstein, D., Hou, W., Miller, T., Mistry, K.B., Swartz, K., 2005. Maternal depressive symptoms and children’s receipt of health care in the first 3 years of life. Pediatrics 115, 306-314.

41. NICHD Early Child Care Research Network, (1999). Chronicity of maternal depressive symptoms, maternal sensitivity, and child functioning at 36 months. Developmental Psychology, 35(5): 1297-1310.

42. NICHD Early Child Care Research Network. (1994). Child care and child development: The NICHD Study of Early Child Care. In S. L. Friedman and H. C. Haywood (Eds.), Developmental follow-up: Concepts, domains, and methods (pp. 377-396). New York: Academic Press.

43. NICHD SECC,www.nichd.nih.gov/research/supported/seccyd/ , Accessed June 2011.

44. Petterson, S.M., Albers A.B., 2001. Effects of poverty and maternal depression on early childhood development. Child Development 72, 1794-1813.

45. Qi, C.H., Kaiser A.P., 2003. Behavior problems of preschool children from low-income families: Review of the literature. Topics in Early Childhood Special Education 23, 188-216.

46. Riggio, H.R., 2006. Introduction: The Adaptive Response of Families to Maternal Employment: Part II – Family Perspectives. American Behavioral Scientist 49, 1303-1309.

47. Ruhm, C.J. (2004). Parental Employment and Child Cognitive Development. Journal of Human Resources, 39, 155-92.

48. Ruhm, C.J. (2008) Maternal employment and adolescent development. Labour Economics, 15, 958-983.

49. Steinberg, L., Sheffield-Morris, A., 2001. Adolescent development. Annual Review of Psychology 52, 83-110.

35

50. Steinberg, L. (2001). We know some things: Parent-adolescent relationships in retrospect and

prospect. Journal of Research on Adolescence, 11(1), 1-19.

51. National Research Council and Institute of Medicine (2000). From Neurons to Neighborhoods: The Science of Early Childhood Development. Committee on Integrating the Science of Early Childhood Development. Jack P. Shonkoff and Deborah A. Phillips, eds. Board on Children, Youth, and Families, Commission on Behavioral and Social Sciences and Education. Washington, D.C.: National Academy Press.

52. US Bureau of Labor Statistics, Employment characteristics of families summary, USDL 09-0568, May 2009.

53. Vandell, D.L., Belsky, J., Burchinal, M., Steinberg, L., Vandergroft, N., NICHD Early Child Care Research Network, 2010. Do Effects of Early Child Care Extend to Age 15 Years? Results From the NICHD Study of Early Child Care and Youth Development. Child Development, 81(3), 737-756.

54. Waldfogel, J., 2002. Child Care, Women’s Employment, and Child Outcomes. Journal of Population Economics 15(3), 527-548.

55. Waldfogel, J., Han, W., Brooks-Gunn, J., 2002. The Effects of Early Maternal Employment on Child Cognitive Development. Demography 39, 369-392.

36

TABLE 1: DESCRIPTIVE STATISTICS, 6 MO. SAMPLE (N = 1,198)

Maternal health and Parenting at 6 months mean SD min max

CES-D score 8.87 8.19 0.50 52.00

Depressed 0.17 0.37 0.00 1.00

Overall health rating (1 = poor, 2 = fair, 3 = good, 4 = excellent) 3.31 0.68 2.00 4.00

Overall health is poor or fair 0.12 0.33 0.00 1.00

Parenting stress score 50.06 9.80 26.00 83.00

Parenting quality score (maternal sensitivity) 9.24 1.78 3.00 12.00

Maternal Employment and Family Income

Mother is employed (either working or on leave) (6 mos.) 0.64 0.48 0.00 1.00

Current weekly hours (6 mos.) 21.21 19.15 0.00 122.00

Average hours in child’s life (1 mo., 3 mo., & 6 mo.) 13.53 12.75 0.00 103.33

Average weekly hours over two prior assessments (1 mo. & 3 mo.) 9.69 11.34 0.00 94.00

Currently works 1 to 20 hours weekly 0.15 0.36 0.00 1.00

Currently works 21 to 39 hours weekly 0.17 0.37 0.00 1.00

Currently works 40+ hours weekly 0.33 0.47 0.00 1.00

Current weekly hours among employed mothers (6 mos.) (n = 772) 32.90 13.56 0.00 122


Average weekly hours two prior assessments among employed (1 mo. & 3 mo.) (n = 772) 14.03 11.34 0.00 94

Currently works 1 to 20 hours weekly, employed mothers (n = 772) 0.23 0.42 0.00 1.00

Currently works 21 to 39 hours weekly, employed mothers (n = 772) 0.26 0.44 0.00 1.00

Currently works 40+ hours weekly, employed mothers (n = 772) 0.51 0.50 0.00 1.00

Mother’s work hours are completely or fairly flexible, employed mothers (n = 772) 0.64 0.48 0.00 1.00

Mother works evening, night, or variable shifts, employed mothers (n = 772) 0.22 0.42 0.00 1.00

Mother’s work involves any overnight travel, employed mothers (n = 772) 0.18 0.39 0.00 1.00

Mother can work from home 10+ hours weekly, employed mothers (n = 772) 0.12 0.33 0.00 1.00

Total family income in dollars, 6 mo. 49340.58 39381.26 2500.00 315000.00

37

Total family income in year prior to birth of child 52391.49 39279.10 2500.00 275001.00

Initial Health & Wellbeing of Family

CES-D score, 1 mo. 11.13 8.85 0.50 53.00

Depressed, 1 mo. 0.24 0.43 0.00 1.00

Overall health rating (1 = poor, 2 = fair, 3 = good, 4 = excellent), 1 mo. 3.48 0.58 2.00 4.00

Overall health is poor or fair (1/0), 1 mo. 0.04 0.20 0.00 1.00

Parenting stress score, 1 mo. 53.14 10.65 27.00 94.00

Health complications during pregnancy 0.32 0.47 0.00 1.00

Child was low birth-weight (<=2500 grams) 0.02 0.15 0.00 1.00

Child was premature (<=37 weeks) 0.04 0.19 0.00 1.00

Mother smoked during pregnancy 0.23 0.40 0.00 1.00

Mother smoked during pregnancy missing 0.08 0.27 0.00 1.00

Maternal Characteristics

Age 28.42 5.54 18.00 46.00

Education in years 14.37 2.47 7.00 21.00

Standardized PPVT score, 36 mos. 99.54 17.15 40.00 159.00

PPVT score missing 0.10 0.30 0.00 1.00

Non-Latino white 0.82 0.38 0.00 1.00

African-American 0.12 0.32 0.00 1.00

Latino 0.04 0.20 0.00 1.00

Other race 0.06 0.25 0.00 1.00

Total household size, 1 mo. 4.05 1.28 2.00 14.00

Married, 1 mo. 0.80 0.40 0.00 1.00

Lives with a partner, 1 mo. 0.08 0.27 0.00 1.00

Single or other family structure, 1 mo. 0.12 0.32 0.00 1.00

Occupation prior to birth of child

Executive, administrative, managerial 0.09 0.28 0.00 1.00

Professional 0.21 0.41 0.00 1.00

Technical or related support 0.04 0.19 0.00 1.00

38

Sales 0.10 0.30 0.00 1.00

Administrative support or clerical 0.22 0.42 0.00 1.00

Private household 0.01 0.10 0.00 1.00

Protective service 0.003 0.05 0.00 1.00

Service 0.11 0.32 0.00 1.00

Farm operation or management 0.003 0.06 0.00 1.00

Mechanic operator, assembler, inspector 0.01 0.07 0.00 1.00

Machine operator, assembler, inspector 0.04 0.19 0.00 1.00

Transportation or material moving 0.003 0.06 0.00 1.00

Handler, equipment cleaner, helper, laborer 0.01 0.10 0.00 1.00

Occupation missing 0.16 0.37 0.00 1.00

Child characteristics

Female child 0.49 0.50 0.00 1.00

Second born 0.35 0.48 0.00 1.00

Third born 0.14 0.35 0.00 1.00

Fourth born or higher birth order 0.06 0.23 0.00 1.00

Child care arrangements at 6 mos.

Child care center, 10+ hours week 0.09 0.29 0.00 1.00

Child care home, 10+ hours week 0.22 0.42 0.00 1.00

Father, 10+ hours per week 0.13 0.34 0.00 1.00

Grandparent, 10+ hours per week 0.11 0.31 0.00 1.00

In-home caregiver, 10+ hours per week 0.09 0.28 0.00 1.00

Multiple arrangements, 10+ hours per week 0.17 0.37 0.00 1.00

Maternal Beliefs

Progressive beliefs about child-rearing, 1 mo. 32.78 3.52 18.00 40.00

Beliefs about benefits of maternal employment, 1 mo. 19.18 3.16 5.00 30.00

Commitment to work, 1 mo. 21.20 5.85 6.00 36.00

39

TABLE 2: MEANS BY WEEKLY WORK HOURS AT 6 MOS., 6 MO. SAMPLE (N = 1,198)

0 HOURS n = 428

1-20 HOURS n = 181

21-39 HOURS n = 198

40+ HOURSn = 391

CES-D score 10.37 7.30*** 7.77*** 8.50***

Depressed 0.22 0.13*** 0.12*** 0.15***

Overall health rating (1 = poor, 2 = fair, 3 = good, 4 = excellent) 3.19 3.44*** 3.37*** 3.33***

Overall health is poor or fair 0.18 0.08*** 0.12** 0.09***

Parenting stress score 51.78 49.31*** 49.63*** 48.74***

Parenting quality score (maternal sensitivity) 9.01 9.75*** 9.39*** 9.18

Current weekly hours (6 mos.) 0.00 13.32*** 31.23*** 42.99***


Average weekly hours over two prior assessments (1 mo. & 3 mo.) 1.92 5.57 12.50 18.68

Total family income in dollars, 6 mo. 37739.50 50870.18*** 56553.04*** 57679.03***

Total family income in year prior to birth of child 45846.98 53549.73** 58863.64*** 55741.69***

CES-D score, 1 mo. 12.76 9.73*** 10.05*** 10.54***

Depressed, 1 mo. 0.31 0.19*** 0.22** 0.21***

Overall health rating (1 = poor, 2 = fair, 3 = good, 4 = excellent), 1 mo. 3.37 3.57*** 3.56*** 3.52***

Overall health is poor or fair (1/0), 1 mo. 0.09 0.01*** 0.03*** 0.02***

Parenting stress score, 1 mo. 54.16 52.59* 53.38** 52.16***

Health complications during pregnancy 0.33 0.32 0.31 0.33

Child was low birth-weight (<=2500 grams) 0.02 0.02 0.03 0.02

Child was premature (<=37 weeks) 0.03 0.03*** 0.04** 0.05*

Mother smoked during pregnancy 0.28 0.16 0.20 0.22

Mother smoked during pregnancy missing 0.09 0.06 0.06 0.10

Age 27.72 29.06*** 28.82** 28.70***

Education in years 13.75 14.93*** 14.82*** 14.56***

Standardized PPVT score, 36 mos. 97.55 103.44*** 100.30* 99.54*

PPVT score missing 0.12 0.07* 0.08 0.10

Non-Latino white 0.76 0.91*** 0.82 0.84***

40

African-American 0.17 0.06*** 0.12* 0.08***

Latino 0.04 0.02 0.04 0.05

Other race 0.07 0.04 0.07 0.07***

Total household size, 1 mo. 4.30 4.14 3.67*** 3.93***

Married, 1 mo. 0.73 0.88 0.83 0.83

Lives with a partner, 1 mo. 0.11 0.05** 0.06** 0.06**

Single or other family structure, 1 mo. 0.16 0.07*** 0.11* 0.10**

Executive, administrative, managerial 0.04 0.06 0.09** 0.15***

Professional 0.11 0.34*** 0.29*** 0.20***

Technical or related support 0.02 0.03 0.07*** 0.04**

Sales 0.08 0.14** 0.12 0.10

Administrative support or clerical 0.16 0.19 0.20 0.32***

Private household 0.01 0.01 0.01 0.01

Protective service 0.00 0.00 0.00 0.01

Service 0.13 0.13 0.13 0.08***

Farm operation or management 0.00 0.01 0.01 0.00

Mechanic operator, assembler, inspector 0.00 0.00 0.01 0.01

Machine operator, assembler, inspector 0.04 0.01** 0.04 0.05

Transportation or material moving 0.01 0.00 0.00 0.00

Handler, equipment cleaner, helper, laborer 0.02 0.00* 0.00* 0.01

Occupation missing 0.38 0.09*** 0.04*** 0.02***

Child care center, 10+ hours week 0.02 0.02 0.15*** 0.17***

Child care home, 10+ hours week 0.06 0.10** 0.32*** 0.41***

Father, 10+ hours per week 0.03 0.19*** 0.18*** 0.20***

Grandparent, 10+ hours per week 0.04 0.09*** 0.22*** 0.14***

In-home caregiver, 10+ hours per week 0.02 0.10*** 0.12*** 0.14***

Multiple arrangements, 10+ hours per week 0.05 0.18*** 0.28*** 0.22***

Progressive beliefs about child-rearing, 1 mo. 32.30 33.09** 33.07*** 33.01***

Beliefs about benefits of maternal employment, 1 mo. 18.32 18.19 19.89*** 20.21***

41

Commitment to work, 1 mo. 20.27 20.06 21.70*** 22.48***

Mother’s work hours are completely or fairly flexible, employed mothers N/A 0.80 0.61 0.58

Mother works evening, night, or variable shifts, employed mothers N/A 0.39 0.23 0.14

Mother’s work involves any overnight travel, employed mothers N/A 0.08 0.21 0.22

Mother can work from home 10+ hours weekly, employed mothers N/A 0.17 0.11 0.11

NOTES: (1) The zero hours category includes 2 respondents who are employed but are on leave at the time of the 6-month interview; (2) T-tests performed on equality of means for each work hour category (1-20, 21-39, 40+) versus the 0 hour category; (3) * denotes statistically significant difference at the .10 level, ** denotes statistically significant difference at the .05 level and *** denotes statistically significant difference at the .01 level; (4) T-tests not conducted for the maternal work characteristics variables since only employed mothers responded to these questions

42

Table 3a: Effect of maternal work hours at 3 mos. on maternal health and parenting measured at 6 mo. wave – Full sample

Model # (1) (2) (3) (4) (5) (6) Dependent Variable:

Log CES-D score 0.003 0.004 0.004 0.004 0.007 0.007 3.95 4.09 3.50 3.08 1.97 1.90Depressed (0/1) 0.0004 0.0006 0.0006 0.0006 0.002 0.002 0.78 1.38 1.48 1.27 2.06 1.96Overall health is poor/fair (0/1) -0.0002 -0.0004 0.0006 -0.0002 -0.002 -0.001 -0.72 -1.04 -0.79 -0.42 -1.88 -1.77Parenting stress score -0.001 0.002 0.002 0.004 0.007 0.009 -0.050 0.02 0.20 0.39 0.51 0.64Parenting quality score -0.001 -0.002 -0.001 -0.002 -0.005 -0.005 -0.23 -0.52 -0.39 -0.50 -1.13 -1.17Covariates: Standard set of covariates X X X X X XMaternal occupation and family income in year prior to childbirth X X X X XFamily structure, maternal work commitment, and maternal beliefs about childrearing and benefits of work, measured at 1 mo.

X X X X

Current family income X XCurrent child care arrangements X X

Notes: (1) Table shows estimates from OLS and probit models (for 0/1 dependent variables) with Huber-White standard errors adjusted for clustering on site; (2) Table shows estimated coefficient (OLS models) or average marginal effect (probit models) and T-stat on maternal weekly work hours measure, measured at 3 mo. telephone interview; (3) Standard set of covariates includes: log CES-D at 1 mo., overall health at 1 mo., parenting stress at 1 mo., maternal age in years, maternal education in years, household size at 1 mo., maternal PPVT reading score, maternal reading score missing (1/0), maternal race/ethnicity (black, other race vs. white, Latino vs. non-Latino), dummy indicators for birth order of child (second, third, fourth or higher vs. firstborn), child was low birthweight (1/0), child was premature (1/0), pregnancy complications (1/0), mother smoked during pregnancy (1/0), mother smoked during pregnancy missing (1/0), dummy indicators for site, and dummy indicators for child birth month; (4) N = 1,198

43

Table 3b: Effect of maternal work hours at 3 mos. on maternal health and parenting measured at 6 mo. wave – Employed sample Model # (1) (2) (3) (4) (5) (6)

Dependent Variable: Log CES-D score 0.006 0.006 0.006 0.006 0.009 0.009 3.43 2.96 2.75 2.51 2.00 1.94Depressed (0/1) 0.001 0.001 0.0004 0.0004 0.002 0.002 1.04 0.85 0.62 0.66 1.75 1.74Overall health is poor/fair (0/1) 0.0001 -0.00003 0.0005 0.001 -0.0005 -0.0005 0.30 -0.09 0.34 0.47 -0.86 -0.81Parenting stress score 0.030 0.027 0.030 0.032 0.029 0.031 3.00 2.66 2.94 3.13 1.90 1.99Parenting quality score -0.002 -0.002 -0.001 -0.001 -0.005 -0.005 -0.42 -0.35 -0.21 -0.26 -0.86 -0.89Covariates: Standard set of covariates X X X X X XMaternal occupation and family income in year prior to childbirth X X X X XFamily structure, maternal work commitment, and maternal beliefs about childrearing and benefits of work, measured at 1 mo.

X X X X

Current family income X XCurrent child care arrangements X X

Notes: (1) Table shows estimates from OLS and probit models (for 0/1 dependent variables) with Huber-White standard errors adjusted for clustering on site; (2) Table shows estimated coefficient (OLS models) and average marginal effect (probit models) and T-stat on maternal weekly work hours measure, measured at 3 mo. telephone interview; (3) Standard set of covariates includes: log CES-D at 1 mo., overall health at 1 mo., parenting stress at 1 mo., maternal age in years, maternal education in years, household size at 1 mo., maternal PPVT reading score, maternal reading score missing (1/0), maternal race/ethnicity (black, other race vs. white, Latino vs. non-Latino), dummy indicators for birth order of child (second, third, fourth or higher vs. firstborn), child was low birthweight (1/0), child was premature (1/0), pregnancy complications (1/0), mother smoked during pregnancy (1/0), mother smoked during pregnancy missing (1/0), dummy indicators for site, and dummy indicators for child birth month; (4) N = 772, mothers from full sample who were employed (working or on leave) at 1 mo. interview.

44

Table 4: Effect of maternal work hours at 3 mos. on maternal overall health measured at 6 mo. wave

PANEL A

Mother’s self-rating of overall health

(3 = excellent, 2 = good, 1= fair or poor)

Ordered Probit Model

Full Sample Employed Sample Work hours at 3 months -0.002

(-1.34) -0.004 (-2.15)

Cut 1 (se)

0.639 (0.800)

0.626 (1.06)

Cut 2 (se)

2.30 (0.816)

2.41 (1.09)

PANEL B Marginal Effects Full Sample Work hours at 3 months

Health is excellent -0.001 (-1.34)

Health is good 0.001 (1.37)

Health is fair or poor 0.0003 (1.28)

Employed Sample Work hours at 3 months

Health is excellent -0.001 (-2.15)

Health is good 0.001 (2.19)

Health is fair or poor 0.0004 (1.99)

Notes: (1) Panel A shows estimated coefficient and T-statistic on maternal work hour measure from an ordered probit model. T-statistics are based on robust standard errors adjusted for clustering on site. (2) Panel B shows marginal effects and T-statistics. Marginal effects indicate the change in the probability of

being in the health category associated with a one hour increase in maternal work hours at 3 months. (3) All models include standard set of covariates described in notes to Table 3 as well as maternal occupation, family income in year prior to childbirth, family structure, maternal work commitment, and maternal beliefs

about childrearing and benefits of work.; (4) Full sample N = 1198; Employed sample N = 772, sample limited to respondents who are employed (either working or on leave) at 1 mo.

45

Table 5: Effects of maternal work hours on maternal health and parenting, 1 mo.- 54 mo. Log CES-D Parenting stress Parenting quality Depressed Poor/fair health Overall health rating Pooled

OLS A-B Pooled

OLS A-B Pooled

OLS A-B Pooled

OLS A-B Pooled

OLS A-B Pooled

OLS A-B

w(t-1) 0.00003 (0.05)

-0.003 (-1.48)

-0.007 (-1.55)

-0.077 (-3.88)

0.460 (20.38)

-0.001 (-0.20)

-0.0004 (-1.64)

-0.001 (-0.49)

-0.0004 (-2.49)

0.00004 (0.08)

0.0004 (1.16)

-0.002 (-2.89)

y(t-1) 0.538 (34.34)

0.028 (0.69)

0.510 (43.89)

0.034 (0.62)

0.003 (1.57)

0.061 (0.75)

0.340 (16.27)

0.080 (2.24)

0.318 (20.62)

0.066 (4.43)

0.474 (43.64)

0.075 (5.86)

n 5,618 4,169 4,573 3,213 4,371 3,071 5,618 4,169 18,655 16,962 18,655 16,962

# instruments 20 12 15 20 240 240 AR(1) test

stat Pr > z

-13.08 (0.00)

-4.28 (0.00)

-6.87 (0.00)

-11.1 (0.00)

-22.48 (0.000)

-26.93 (0.00)

AR(2) test stat

Pr > z

1.02 (0.307)

-2.43 (0.015)

-0.05 (0.959)

0.38 (0.706)

-0.46 (0.649)

0.45 (0.655)

Overiden. test

19.79 (0.071)

15.48 (0.009)

8.49 (0.387)

14.65 (0.261)

233.29 (0.272)

211.54 (0.665)

Notes: (1) Table 5 only shows estimated coefficients and T-statistics on lagged work hours and lagged dependent variable. Models also include: whether father is in household, household size, and dummy variables for survey wave – estimated coefficients not shown; (2) In pooled OLS models, robust standard errors adjusted for clustering on respondent are shown; (3) Arellano-Bound estimates generated using one-step, difference A-B estimator with robust standard errors.

46

Appendix Table 1: Effect of maternal work hours on maternal health and parenting measured at 6 month SECC interview OLS Models with alternate measures of maternal employment

Dependent Variable Log CES-D score Depressed Overall health poor/fair Parenting stress score Parenting quality score

Panel A: Full Sample Ave hours in child’s life 0.004 0.0004 -0.001 -0.012 -0.003 2.96 0.77 -1.28 -0.54 -0.58 Ave hours in prior 2 waves 0.005 0.001 -0.001 0.016 -0.001 2.35 1.33 -1.04 0.75 -0.30 1-20 hours, 3 mo. 0.028 -0.024 -0.021 0.35 0.05 0.30 -0.43 -0.66 0.44 0.92 21-39 hours, 3 mo. 0.027 -0.022 -0.004 0.53 0.010 0.25 -0.74 -0.11 0.74 0.06 40+ hours, 3 mo. 0.156 0.013 -0.021 -0.025 -0.027 3.86 0.75 -1.11 -0.06 -0.14

Panel B: Employed Sample Ave hours in child’s life 0.007 0.0001 -0.002 0.037 -0.003 2.03 0.62 -0.47 1.63 -0.57 Ave hours in prior 2 waves 0.007 0.0001 -0.0005 0.060 -0.0003 1.76 0.86 -1.05 3.17 -0.1 1-20 hours, 3mo. 0.05 -0.016 -0.026 1.47 -0.005 0.48 -0.35 -1.06 2.37 -0.05 21-30 hours, 3 mo. 0.07 -0.048 0.02 1.60 0.012 0.62 -1.50 0.95 2.60 0.05 40+ hours, 3 mo. 0.22 -0.001 -0.009 1.34 -0.035 2.65 -0.05 -0.76 2.90 -0.14

Notes: (1) Table shows estimated coefficients and T-statistics in parentheses from OLS models with Huber-White standard errors adjusted for clustering on site; (2) Table shows estimated coefficient on maternal work hours measure(s) only. Work hours measures are: average

hours per week worked in child’s life; Average hours per week in prior two interviews (1 and 3 mos.); Hours worked in prior (3 mo.) interview, 1-20 hrs/wk; Hours worked in prior (3 mo.) interview, 21-39 hrs/wk; Hours worked in prior (3 mo.) interview, 40+ hrs/wk (3) All models include standard set of covariates described in notes to Tables 3a-b as well as maternal occupation, family income in year prior to childbirth, family structure, maternal work commitment, and

maternal beliefs about childrearing and benefits of work.; (4) Full sample N = 1198; Employed sample N = 772, sample limited to respondents who are employed (either working or on leave) at 1 mo.

47

Appendix Table 2: Effect of work characteristics on maternal health and parenting measured at 6 month SECC interview –Employed sample

Log CES-D score Depressed (1/0) Poor/fair health (1/0) Parenting stress index Parenting quality score Hours worked per week in prior (3 mo.) interview 0.005 0.000 0.000 0.030 0.000 2.450 0.460 0.020 2.840 -0.220 1-20 hours worked in prior interview -0.019 -0.022 -0.014 1.152 -0.021 -0.170 -0.500 -0.520 1.890 -0.370 21-39 hours worked in prior interview 0.051 -0.053 0.026 1.473 0.050 0.430 -2.030 1.150 2.470 0.610 40+ hours worked in prior interview 0.202 -0.008 -0.013 1.267 -0.031 2.430 -0.350 -0.810 2.470 -0.340 Flexible hours -0.014 -0.010 0.008 0.004 -0.028 -0.027 -0.204 -0.272 0.032 0.034 -0.160 -0.110 0.260 0.140 -1.040 -1.010 -0.410 -0.540 0.590 0.580 Overnight travel 0.020 0.022 -0.006 -0.005 -0.007 -0.007 0.125 0.233 -0.029 -0.028 0.180 0.210 -0.180 -0.170 -0.260 -0.260 0.200 0.360 -0.500 -0.470 Non-standard hours 0.164 0.178 -0.006 0.000 0.003 0.003 1.498 1.331 0.037 0.039 2.270 2.150 -0.170 -0.010 0.140 0.150 3.340 2.860 0.710 0.700 Works from home 10+ hrs 0.014 0.016 -0.002 -0.006 -0.066 -0.064 -0.512 -0.489 0.035 0.040 0.140 0.150 -0.070 -0.190 -2.630 -2.400 -0.720 -0.700 0.460 0.530

Notes: (1) For continuous variables, table shows estimated coefficients and T-statistics in parentheses from OLS models with Huber-White standard errors adjusted for clustering on site; for dichotomous variables, average marginal effect is shown instead of coefficient (2) Table shows estimated coefficient on maternal work hours and job characteristics measure(s) only. (3) All models include standard set of covariates described in notes to Tables 3a-b as well as maternal occupation, family income in year prior to childbirth, family structure, maternal work commitment, and maternal beliefs about childrearing and benefits of work.; (4) N = 772, sample limited to respondents who are employed (either working or on leave) at 1 mo.

early maternal employment and family wellbeing

Documents